Leveraging machine learning to examine engagement with a digital therapeutic.

Digital Therapeutics (DTx) are evidence-based software-driven interventions for the prevention, management, and treatment of medical disorders or diseases. DTx offer the unique ability to capture rich objective data about when and how a patient engages with a treatment. Not only can one measure the quantity of patient interactions with a digital treatment with high temporal precision, but one can also assess the quality of these interactions. This is particularly useful for treatments such as cognitive interventions, where the specific manner in which a patient engages may impact likelihood of treatment success. Here, we present a technique for measuring the quality of user interactions with a digital treatment in near-real time. This approach produces evaluations at the level of a roughly four-minute gameplay session (mission). Each mission required users to engage in adaptive and personalized multitasking training. The training included simultaneous presentation of a sensory-motor navigation task and a perceptual discrimination task. We trained a machine learning model to classify user interactions with the digital treatment to determine if they were "using it as intended" or "not using it as intended" based on labeled data created by subject matter experts (SME). On a held-out test set, the classifier was able to reliably predict the SME-derived labels (Accuracy = .94; F1 Score = .94). We discuss the value of this approach and highlight exciting future directions for shared decision-making and communication between caregivers, patients and healthcare providers. Additionally, the output of this technique can be useful for clinical trials and personalized intervention.

A novel digital intervention for actively reducing severity of paediatric ADHD (STARS-ADHD): a randomised controlled trial.

BACKGROUND: Attention-deficit hyperactivity disorder (ADHD) is a common paediatric neurodevelopmental disorder with substantial effect on families and society. Alternatives to traditional care, including novel digital therapeutics, have shown promise to remediate cognitive deficits associated with this disorder and may address barriers to standard therapies, such as pharmacological interventions and behavioural therapy. AKL-T01 is an investigational digital therapeutic designed to target attention and cognitive control delivered through a video game-like interface via at-home play for 25 min per day, 5 days per week for 4 weeks. This study aimed to assess whether AKL-T01 improved attentional performance in paediatric patients with ADHD. METHODS: The Software Treatment for Actively Reducing Severity of ADHD (STARS-ADHD) was a randomised, double-blind, parallel-group, controlled trial of paediatric patients (aged 8-12 years, without disorder-related medications) with confirmed ADHD and Test of Variables of Attention (TOVA) Attention Performance Index (API) scores of -1·8 and below done by 20 research institutions in the USA. Patients were randomly assigned 1:1 to AKL-T01 or a digital control intervention. The primary outcome was mean change in TOVA API from pre-intervention to post-intervention. Safety, tolerability, and compliance were also assessed. Analyses were done in the intention-to-treat population. This trial is registered with ClinicalTrials.gov, NCT02674633 and is completed. FINDINGS: Between July 15, 2016, and Nov 30, 2017, 857 patients were evaluated and 348 were randomly assigned to receive AKL-T01 or control. Among patients who received AKL-T01 (n=180 [52%]; mean [SD] age, 9·7 [1·3] years) or control (n=168 [48%]; mean [SD] age, 9·6 [1·3] years), the non-parametric estimate of the population median change from baseline TOVA API was 0·88 (95% CI 0·24-1·49; p=0·0060). The mean (SD) change from baseline on the TOVA API was 0·93 (3·15) in the AKL-T01 group and 0·03 (3·16) in the control group. There were no serious adverse events or discontinuations. Treatment-related adverse events were mild and included frustration (5 [3%] of 180) and headache (3 [2%] of 180). Patient compliance was a mean of 83 (83%) of 100 expected sessions played (SD, 29·2 sessions). INTERPRETATION: Although future research is needed for this digital intervention, this study provides evidence that AKL-T01 might be used to improve objectively measured inattention in paediatric patients with ADHD, while presenting minimal adverse events. FUNDING: Sponsored by Akili Interactive Labs.

Aging, Spatial Disparity, and the Sound-Induced Flash Illusion.

The present study examined age-related differences in multisensory integration and the effect of spatial disparity on the sound-induced flash illusion--an illusion used in previous research to assess age-related differences in multisensory integration. Prior to participation in the study, both younger and older participants demonstrated their ability to detect 1-2 visual flashes and 1-2 auditory beep presented unimodally. After passing the pre-test, participants were then presented 1-2 flashes paired with 0-2 beeps that originated from one of five speakers positioned equidistantly 100 cm from the participant. One speaker was positioned directly below the screen, two speakers were positioned 50 cm to the left and right from the center of the screen, and two more speakers positioned to the left and right 100 cm from the center of the screen. Participants were told to report the number of flashes presented and to ignore the beeps. Both age groups showed a significant effect of the beeps on the perceived number of flashes. However, neither younger nor older individuals showed any significant effect of spatial disparity on the sound-induced flash illusion. The presence of a congruent number of beeps increased accuracy for both older and younger individuals. Reaction time data was also analyzed. As expected, older individuals showed significantly longer reaction times when compared to younger individuals. In addition, both older and younger individuals showed a significant increase in reaction time for fusion trials, where two flashes and one beep are perceived as a single flash, as compared to congruent single flash trials. This increase in reaction time was not found for fission trials, where one flash and two beeps were perceived as two flashes. This suggests that processing may differ for the two forms for fission as compared to fusion illusions.

Behavioral training to improve collision detection.

Young drivers are a high-risk group for vehicle crashes due to inexperience in detecting an impending collision and are one group that may benefit from perceptual learning (PL) training. The present study assessed whether PL could be used to improve performance in collision detection. Ten college-aged subjects participated in the first experiment, which consisted of seven 1-hr sessions conducted on separate days. Thresholds at three observer/object speeds were measured prior to training using a two-alternative forced choice procedure during which they indicated whether an approaching object would result in a collision or noncollision event. Participants were then trained near threshold at one of these speeds for 5 days. After training, participants showed a significant reduction in the time needed to detect a collision at the trained speed. This improvement was also found to transfer to the higher observer speed condition. A second experiment was conducted to determine whether this improvement was due to training near threshold or whether this improvement was merely due to practice with the task. Training with stimuli well above threshold showed no significant improvement in performance, indicating that the improvement seen in the first experiment was not solely due to task practice.

Improving vision among older adults: behavioral training to improve sight.

A major problem for the rapidly growing population of older adults (age 65 and over) is age-related declines in vision, which have been associated with increased risk of falls and vehicle crashes. Research suggests that this increased risk is associated with declines in contrast sensitivity and visual acuity. We examined whether a perceptual-learning task could be used to improve age-related declines in contrast sensitivity. Older and younger adults were trained over 7 days using a forced-choice orientation-discrimination task with stimuli that varied in contrast with multiple levels of additive noise. Older adults performed as well after training as did college-age younger adults prior to training. Improvements transferred to performance for an untrained stimulus orientation and were not associated with changes in retinal illuminance. Improvements in far acuity in younger adults and in near acuity in older adults were also found. These findings indicate that behavioral interventions can greatly improve visual performance for older adults.

Optimization of perceptual learning: effects of task difficulty and external noise in older adults.

Previous research has shown a wide array of age-related declines in vision. The current study examined the effects of perceptual learning (PL), external noise, and task difficulty in fine orientation discrimination with older individuals (mean age 71.73, range 65-91). Thirty-two older subjects participated in seven 1.5-h sessions conducted on separate days over a three-week period. A two-alternative forced choice procedure was used in discriminating the orientation of Gabor patches. Four training groups were examined in which the standard orientations for training were either easy or difficult and included either external noise (additive Gaussian noise) or no external noise. In addition, the transfer to an untrained orientation and noise levels were examined. An analysis of the four groups prior to training indicated no significant differences between the groups. An analysis of the change in performance post-training indicated that the degree of learning was related to task difficulty and the presence of external noise during training. In addition, measurements of pupil diameter indicated that changes in orientation discrimination were not associated with changes in retinal illuminance. These results suggest that task difficulty and training in noise are factors important for optimizing the effects of training among older individuals.

Multisensory integration, aging, and the sound-induced flash illusion.

The present study examined age-related differences in multisensory integration and the role of attention in age-related differences in multisensory integration. The sound-induced flash illusion--the misperception of the number of visual flashes due to the simultaneous presentation of a different number of auditory beeps--was used to examine the strength of multisensory integration in older and younger observers. The effects of integration were examined when discriminating 1-3 flashes, 1-3 beeps, or 1-3 flashes presented with 1-3 beeps. Stimulus conditions were blocked according to these conditions with baseline (unisensory) performance assessed during the multisensory block. Older participants demonstrated greater multisensory integration--a greater influence of the beeps when judging the number of visual flashes--than younger observers. In a second experiment, the role of attention was assessed using a go/no-go paradigm. The results of Experiment 2 replicated those of Experiment 1. In addition, the strength of the illusion was modulated by the sensory domain of the go/no-go task, though this did not differ by age group. In the visual go/no-go task we found a decrease in the illusion, yet in the auditory go/no-go task we found an increase in the illusion. These results demonstrate that older individuals exhibit increased multisensory integration compared with younger individuals. Attention was also found to modulate the strength of the sound-induced flash illusion. However, the results also suggest that attention was not likely to be a factor in the age-related differences in multisensory integration.